Munger: “You must know the big ideas in the big disciplines and use them routinely — all of them, not just a few..... “You've got to have models in your head, and you've got to array your experience — both vicarious and direct — on this latticework of models.” (2)

Sutton’s post got me thinking about Munger’s latticework of mental models and my experience how AI is already moving in that direction.

One practical application is PARTNERSHIP: I bring an investment thesis and ask AI to examine it through one or two additional lenses —I inquire specifically: investor psychology, inversion, disconfirming evidence, or whatever models might be relevant. AI even identifies variables I’ve overlooked—importantly: I decide whether they matter: investigate the important ones and ignore the noise.

We’ve already been doing this, whether formally labeled or not. In one recent analysis of a company’s debt, no single concern was necessarily a deal killer, but several variables, taken together, moved the investment into my “too tough” pile. THAT ALONE is HUGE. AI convinced me: this wasn’t the fastball down the middle I thought it was, it was a hard breaking curve on the outside corner. I didn’t swing.

The beauty is neither partner has to be infallible. If AI surfaces just one or two valuable insights I hadn’t considered, the exercise may be worthwhile. And sometimes the most valuable insight is recognizing an opportunity is simply too difficult to understand confidently.

The goal isn’t to have AI make our decisions. It’s to use AI to EXPAND our latticework, challenge our assumptions, and improve our judgment. Even if it’s just at the margin. It already showed me holes in my theses. It’s already saved me from potentially bad decisions.

As Munger might put it, we’re trying to make fewer mistakes — not swing at more pitches.